Discovering the transporters used by specific drugs can have profound impacts on patient care, and can also inform drug development. Drugs taken orally must pass through the digestive tract, and this often happens via transporter proteins. But, it’s often unknown which transporter a certain drug uses to exit the digestive tract, and this could potentially lead to drug interference. A team of researchers at MIT, Brigham and Women’s Hospital, and Duke University is now changing this.
The researchers have developed a unique method to identify the transporters for different drugs using tissue models and machine-learning algorithms. Their tests have revealed that an antibiotic commonly given to patients can interfere with a blood thinner. The team highlights that understanding which transporters help drugs move through the digestive tract can help drug makers improve drug absorbability, by including excipients that help increase their interaction with transporters.
The team’s study is the first to systematically evaluate the role of the three most used GI transporters – BCRP, MRP2, PgP – in drug transport. The researchers analyzed the drugs’ efficacy using pig intestinal tissue in the lab and examined how transporters interact with over twenty commonly used drugs. Using a machine-learning model trained on the resultant data, the researchers could draw up almost 2 million predictions of potential drug interactions.
These predictions included the interactions of doxycycline, an antibiotic, with drugs like warfarin, digoxin, levetiracetam, and tacrolimus. They used patient data from Massachusetts General Hospital and Brigham and Women’s Hospital and found that the blood levels of the drugs fluctuated due to interactions with doxycycline.
The team emphasizes that their approach cannot only help identify interactions between drugs already in use but can also aid in formulating new molecules by reducing interactions with other drugs and increasing their absorbability.
Vivtex, a biotech company co-founded by MIT professor Robert Langer and the senior author of the study, Associate Professor Giovanni Traverso, is using this concept to develop new oral drug delivery systems. Being the first to predict and validate these drug interactions using their novel model, the team believes they have developed a method that will allow us to better understand the safety implications of administering certain drugs together.
The research was partially funded by the U.S. National Institutes of Health, the Department of Mechanical Engineering at MIT, and Brigham and Women’s Hospital. Other key members of the research team include former MIT postdocs Yunhua Shi and Daniel Reker. The study has been published in Nature Biomedical Engineering.
